1

Microbial hydrogen-sulphide elimination in continuous biotrickling reactor by

immobilized Thiobacillus thioparus

Gábor Tóth, Éva Lövitusz, Nándor Nemestóthy, Katalin Bélafi-Bakó*

University of Pannonia, Research Institute on Bioengineering, Membrane

Technologies and Energetics

10. Egyetem Str., Veszprém, 8200 Hungary

Tel. +36 (88) 624726, Fax: +36 (88) 624292

E-mail: bako@almos.uni-pannon.hu

Abstract

Nowadays removal of hydrogen sulphide from gaseous streams by biological

treatments is a promising alternative procedure, among them biotrickling reactor

seems a reliable and efficient system. To maximize the performance, strains should

have high hydrogen sulphide elimination efficiency; excellent carriers should be

selected where the microbes can be immobilized. In this study various carriers were

used as the support medium for the immobilization of Thiobacillus thioparus and a

continuous biotrickling reactor was constructed and operated for H2S removal. We

found that our systems with Mavicell and Kaldnes supports are able to remove H2S

from the gas mixture with high efficiency (95-100 %), and the specific removal

capacity was calculated as a high as 30-40 g S/m3h.

2

1. Introduction

Biological techniques for hydrogen sulphide elimination can be applied in wide

range and seem quite promising in certain circumstances where the aim is to remove

„smelly” compounds. Among these components hydrogen sulphide is one of the

most important substances, since its smelling limit value is rather low, 0.5-2.0 ppb

[1].

Nowadays the biological removal of air pollutants has been studied intensively [2].

One of the intensification methods of these biosystems is the immobilization of the

microbes in a form of biofilm, which exploits the natural bounding capability of

certain microorganisms on a given surface, thus the pollutants can be eliminated with

higher effectiveness [3]. The performance of an immobilized film bioreactor can be

enhanced by selection of a proper support material for the given microorganism [4].

The suitable supports provide optimal conditions for the microbes, having high

specific surface area [5].

Currently natural support materials including soil, compost, peat ...etc. are often used

as media for biofiltration [6]. Although these materials are considered as rather cost

effective media, their practical application is still limited [7] mainly due to their

aging which causes declining effectiveness. Therefore the novel research tendency

leads towards the synthetic materials [8-10], including ceramic saddles polyethylene

pall rings, synthetic foams, activated carbon, extruded diatomaceous earth pellets,

glass beads and Ca-alginate.

3

In our experiments Mavicell-B cellulose beads, activated carbon, polyethylene rings

(Kaldnes-K1) and alginate beads were used as support materials. All of these

supports are synthetic materials, two of them (activated carbon and alginate) have

already been studied, but no literature reports on application of Kaldnes-K1 and

Mavicell-B have been found so far.

The microbes can be grown onto the surface of the activated carbon, Mavicell and

Kaldnes, thus the affinity of the bacteria to the surface area influenced highly the

quality and thickness of the biofilm. In case of alginate, however, microbes are

entrapped into the alginate beads (known amount of bacteria), thus it does not

depend on the surface of the support. Hence the two different immobilization

methods can be compared.

Bacteria belonging to the Thiobacillus strains have higher hydrogen sulphide

elimination efficiency than other sulphide oxidising microorganisms [9, 10]. In our

preliminary experiments two colourless sulphur bacteria were studied in a batch

system (Thiomonas intermedia, Thiobacillus thioparus). They were immobilized on

three different supports and the operational stabilities were compared [12]. The

results have shown that the degradation ability of immobilised Thiobacillus

thioparus was higher both in soluble and immobilised forms. Therefore the

experiments were continued with this bacteria aiming to construct a biotrickling

reactor and accomplish a successful continuous system for hydrogen sulphide

elimination from gas streams.

2. Materials and methods

4

2.1. Microorganism

The strain Thiobacillus thioparus was purchased from the strain collection of DSMZ

(Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, Germany). It

was grown on a special Thiomonas intermedia broth, its composition is as follows

(g/L): NH4Cl 0.1, KH2PO4 3.0, MgCl2*6 H2O 0.1, CaCl2 0.1, Na2S2O3*5 H2O 5.0,

yeast extract 1.0, and 1000 ml distilled water [13]. 33 °C temperature and 120 rpm

shaking were maintained for incubation under sterile condition. The bacterium used

sulphide as an energy source in the multistep oxidation procedure, thus oxidised

sulphur compounds are formed, causing acidification of the system [15][16]. To

avoid it phosphate buffer was applied to maintain pH at 5.8 (adding 0,356 g K2HPO4

to every l broth).

2.2. Immobilisation of the bacteria

50 ml concentrated inocula (its total solid substance, TSS was 0.38 g/L) and 140 ml

sterile broth were added to 70 ml sterilized support and it was incubated for 2-3 days.

The immobilization was followed by protein determination. The bacteria

immobilized on the support was filled into a glass column, thus the experiments were

carried out from this point under non-sterile conditions. A “blind” column was used

for comparison purposes, its infection was prevented by using 1.5 % sodium

benzoate solution.

2.3. Supports used

5

The first support used was alginate beads, widely used in biotechnology for

immobilization of cells and enzymes (Figure 1) by the so called entrapment

technique. During jellification small hollows are formed in alginate where

biocatalysts (enzymes and cells) can be entrapped. The structure of the gel is

compatible with the biocatalysts, thus no chemical modification is needed [16, 17].

The second support, the granulated activated carbon (GAC) was purchased from

Airwatec s.a. (Belgium) (Figure 2). Due to its high surface area it seems also a

promising support material for immobilisation of microorganisms [18]. The features

are listed in Table 1.

MAVICELL-B (Table 2) is cellulose beads, purchased from Magyar Viscosagyár,

(Nyergesújfalu, Hungary) widely used for immobilization of various microbes,

having large adsorption surface area, thus a highly suitable support, moreover it can

withstand to the corrosive effect of hydrogen sulphide. The cells can be bound on the

surface of the support by adsorption.

Finally Kaldnes K1 polyethylene rings (Evolution Aqua, Lancashire, UK) were used

as supports (Figure 3), which is often applied in waste water treatment technologies,

where biofilms are needed. its length is 7 mm, diameter 10 mm. Since the diameter

of the column planned to use as reactor is similar, the rings were splitted into two

halves.

2.4. Designing the continuous reactor system

6

Experiments were carried out in two parallel, same volume columns (Table 3 and

Figure 4). The columns can be divided into three parts: the first one is a thermostated

(jacketed) reactor space containing the bacteria immobilized on the support; the

second part is chilled by a cryostat to remove the water vapour by condensation,

while the third part is a thermostated one again. The feed stream (model gas to be

separated) is introduced in the bottom of the first part of the column. The hydrogen

sulphide concentration was followed by a gas sensor, placed on the top of the column

reactors. It is a sensitive sensor, thus it is important to maintain the temperature of

the gas stream in the same level, moreover to remove moisture, that’s why were the

two upper parts built in the system.

Similar gas mixture (same inlet rate: 360 ml/min and H2S concentration: 80-100

ppm) was introduced to both columns, while a broth was trickling (rate 0.7 ml/min)

onto the support packed in the column, which was collected in the bottom of the

column and recirculated by a peristaltic pump.

The column was characterized by numerous data, summarized in Table 3. The

average reaction time was 200-220 hours.

2.5. Gas mixture used

In the experiments a model gas mixture was used containing 40-44% (v/v) CO2, 1-2

% (v/v) O2, 80-100 ppm H2S and 54-58 % (v/v) N2.

2.6. Analysis

7

2.6.1. Following the gas composition

The gas was introduced in the bottom of the column and went through the active

support, thus its H2S concentration decreased. This reduction was followed by a

FIGARO TGS 825 sensor placed on the top of the column. These TGS (Taguchi Gas

Sensor) type sensors are based on a metal oxide, which are contaminated by some

noble metal. These metals – upon heating – can react with the de-oxidising gases

present (e.g.: H2S), which gives a signal in the resistance of the cell. The higher the

gas concentration, the lower is the resistance [19]. The measuring range of the sensor

is 0-100 ppm and it was calibrated by a Drager X-am 7000 type mobile gas analyzer.

2.6.2. Protein determination

Protein content was determined by the modified Folin method, which gives a blue

colour reaction with proteins in alkali media. The reaction took 30 min and the

solutions were measured at 720 nm. Calibration was carried out by using BSA

protein.

3. Results

3.1. Alginate

As a result of the immobilization procedure, Thiobacillus thioparus bacteria were

successfully entrapped in the alginate beads, and finally 6.5-7 mg protein / g support

8

was measured. Alginate beads with the bacteria were packed into the column and

experiments were carried out to study the H2S reduction in the gas stream. However,

the H2S level did not decline, though the protein level has not decreased, indicating

that the microbes were still there. It seemed that the structure of the beads has

changed, somehow they have lost their water content and the mechanical stability.

The volume of the beads decreased and the support was shrinking due to its own

weight, losing the majority of the surface area. That’s why it did not work properly.

Chang and co-authors have carried out similar experiments [20] with alginate where

Thiobacillus thioparus was entrapped. They saturated the gas stream with water

vapour before introducing it, thus the support packed did not lose its water content

and the stability was maintained during operation. In industrial applications,

however, the aim is to prepare a stabile, easy-to-handle and mechanically strong

support and alginate does not seem sturdy enough here.

3.2. Activated carbon

The next support was activated carbon, where the bacteria were immobilized. The

results of the experiments in the column are shown in Figure 5. As it can be seen the

H2S was not eliminated properly, the H2S content in the gas fluctuated randomly

(Figure 5a), the operation was not satisfactory, though the presence of the microbes

were proven by checking the protein content (Figure 5b). It seemed again that the

structure of the support caused the problem. The activated carbon granules were

sticking together, they formed plugs (Figure 6) which hindered gas flowing in the

9

column. These plugs started to rise up slowly, up to the sensor. Thus we found that

the support was not suitable for H2S elimination from the gas.

3.3. Mavicell

Our next support was Mavicell where the bacteria were immobilized. The beads were

packed to the reactor and experiments were carried out with circulating the model

gas mixture. The results are presented in Figure 7. Starting the gas introduction into

the column, the H2S concentration in the gas stream leaving the trickling bioreactor

was declining sharply, and finally it was stabilized at the level of 5 ppm. It means

that in the reactor 90-95 % of the H2S was removed compared to the control column

where no microbes were present. The bacteria on the Mavicell support worked

effectively, their amount was slightly increased according to the data on protein

determination (Figure 7b).

During the steady state operation period the specific removal capacity (productivity)

of the column (related to the reactor volume) was calculated as 30 g H2S/m3h, which

is a similar value as Oyarzu´n et al. reported [21] using Thiobacillus thioparus

immobilized on peat, while Ramírez et al [22] achieved lower specific capacity (14.9

g S/m3h) using polyurethane foam as support, with similar removal efficiency (99,8

%).

3.4. Kaldnes K1 media

10

In case of Kaldens K1 support the experimental results were similar to the Mavicell

(Figure 8). H2S concentration in the gas was reduced by the bacteria from the initial

90 pm down to 0-5 ppm, which means 95-100% removal efficiency. The amount of

microbes on the Kaldnes support was 20 % higher than in Mavicell beads (Figure

8b).

The specific removal capacity during the steady state operation was calculated as 35-

40 g H2S/m3h, which is slightly higher than it was in Mavicell. This value is even

higher than the result of Eliasa et al. [23], who used a mixture of pig manure and

sawdust as support, and reported 28.5 g H2S/m3h specific capacity with >95%

removal efficiency and 40.5 H2S/m3h specific capacity with >90% removal

efficiency.

4. Conclusions

A continuous biotrickling column reactor was designed and operated packed with

Thiobacillus thioparus bacteria immobilized on various supports (Alginate, activated

carbon, Mavicell and Kaldnes) for H2S elimination from gaseous streams.

Application of Mavicell cellulose beads and Kaldnes K1 polyethylene rings as

supports for this colourless sulphur oxidising bacteria has not been reported so far,

hence the results obtained were compared to other supports

Our experiments have proven that Thiobacillus thioparus bacteria can be

immobilized onto these supports, among them Kaldnes K1 was found the best. We

believe that our systems with Mavicell and Kaldnes supports are able to remove H2S

from the gas mixture with high efficiency (95-100 %), and the specific removal

11

capacity was calculated 30-40 g S/m3h, which is similar to the literature data.

Therefore we think that these biotrickling systems are suitable for H2S elimination

from gases. Now the aim is to carry out long-term experiments (operation stability,

reliability) using controlled oxygen concentration. Although oxygen should be

present in these systems (since the bacteria need it), but its high level is not desirable

(methane – oxygen mixture!), therefore in the next set of experiments we will try to

lower its level to reach a minimum value where the systems still work properly.

References

[1] Stuetz, R., Frechen, F. B., Odours in Wastewater Treatment: Measuring,

Modelling and Control, Int. Water Assoc., London 2001.

[2] Gabriel, D., Deshusses, M.A., Retrofitting Existing Chemical Scrubbers to

Biotrickling Filters for H2S Emission Control, Proc. Nat. Acad. Sci. U.S.A, 2003,

100 , 11, 6308-6312.

[3] Rattanapan, C., Boonsawang, P., Kantachote, D., Removal of H2S in down-

flow GAC biofiltration using sulfide oxidizing bacteria from concentrated latex

wastewater, Bioresource Technology, 2009, 100, 125–130.

[4] Luo, J., A pilot-scale study on biofilters for controlling animal rendering

process odours, J. Water Sci. Technol, 2001, 44, (9), 277–285.

[5] Elias, A., Barona, A., Arreguy, A., Rios, J., Aranguiz, I., Penas, J.,.

Evaluation of a packing material for the biodegradationof H2S and product analysis,

Process Biochem, 2002, 37, 813–820.

12

[6] Leson, G., M. Winer, A., Biofiltration: An Innovative Air Pollution Control

Technology For VOC Emissions, Journal of the Air & Waste Management

Association, 2012, 41, 8, 1045-1054.

[7] Van Langenhove, H., Bendinger, B., Oberthür, R., Schamp, W. Organic

sulfur compounds: persistant odourants in the biological treatment of complex

waste gases [In:] Dragt, A. J., Van Dam, J., Eds. Biotechniques for Air Pollution

Abatement and Odour Control Policies, Elsevier, Amsterdam, 1992, 309–313.

[8] Cho, K.S., Ryu, H.W., Lee, N.Y., Biological deodourization of hydrogen

sulphide using porous lava as a carrier of Thiobacillus thiooxidans, J. Biosci.

Bioeng, 2000, 90, 1, 25–31.

[9] Atlas, JE., Bartha, R., Microbial Ecology: Fundamentals and Application,

Addison-Wesley Publishing Company, Boston, 2001, USA,

[10] Rattanapan, C., Ounsaneha W., Removal of Hydrogen Sulfide Gas using

Biofiltration - a Review, Walailak J Sci & Tech, 2012 9, 1, 9‐18.

[11] Winogradsky, S., Beitrage zur Morphologie und Physiologie der Bakterien.

Heft 1. Zur Morphologie und Physiologie der Schwefelbakterien. 1:1201888

[12] Tóth, G., Lövitusz, É., Nemestóthy, N., Bélafi-Bakó, K., Biocatalytic

Hydrogen-sulphide Removal from Gaseous Streams, Hungarian Journal of

Industrial Chemistry, 2013, 41, 1, 23-27.

[13] www.dsmz.de

[14] Robertson, L. A., Kuenen, J.G., The Colorless Sulfur Bacteria [In:]

Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K-H., Stackebrandt, E., The

Prokaryotes, A Handbook of the Biology of Bacteria, Springer science+Business

Media, New York, 2006, 2, 985–1011

13

[15] Tang, K., Baskaran, V., Nemati, M., Bacteria of the sulphur cycle: An

overview of microbiology, biokinetics and their role in petroleum and mining

industries, Biochemical Engineering Journal, 2009, 44, 73–94.

[16] Boross, L., Sisak, Cs., Szajáni, B., Solid phase biocatalists – Preparation,

features, practical applications, Akadémiai Kiadó, Budapest, (in Hungarian) 2008,

53-54.

[17] Sevella, B., Bioengineering processes, Typotex, Budapest (in Hungarian)

2012.

[18] Airwatec S.A. Belgium

[19] www.figarosensor.com/

[20] Chung, Y-C., Huang Ch., Tsengb, C.P., Operation optimization of

Thiobacillus thioparus CH l1 biofilter for hydrogen sulfide removal, Journal of

Biotechnology, 1996, 52, 31-38.

[21] Oyarzun, P., Arancibia, F., Canales, C., Aroca, G. E., Biofiltration of high

concentration of hydrogen sulphide using Thiobacillus thioparus, Process

Biochemistry, 2003, 39, 165-170.

[22] Ramírez, M., Gómez, J. M., Aroca, G., Cantero, D., Removal of hydrogen

sulfide by immobilized Thiobacillus thioparus in a biotrickling filter packed with

polyurethane foam, Bioresource Technology, 2009, 100 4989–4995.

[23] Eliasa, A., Barona, A., Arreguy, A., Rios, J., Aranguiz, I., Penas J.,

Evaluation of a packing material for the biodegradation of H2S and product analysis,

Process Biochemistry, 2002, 37, 813 – 820.

[24] Cho, K-S., Hirai, M., Shoda, M., Enhanced Removal Efficiency of

Malodorous Gases in a Pilot-Scale Peat Biofilter Inoculated with Thiobacillus

thioparus DW44, Journal of fermentation and bioengineering, 1992, 73, 1, 46-50.

14

Figure 1: Alginate beads

Figure 2: Activated carbon granules used

Table 1: The parameters of activated carbon

Parameter Value

Total surface area (BET) (m2/g) 1080

pH 7

Water content (%) 1,1

Ash content (%) 8,6

Granules Diameter (mm) 1

15

Table 2:Features of MAVICELL-B

Figure 3: Kaldnes K1 polyethylene rings

Table 1: Features of the bioreactor

Feature Value

Height of each column (mm) 250

Diameter of each column (mm) 20

Volume of each media (ml) 70

Porisity of Mavicell-B (ml) 25

Gas retention time in Mavicell-B (s) 4,1

Porisity of activated carbon (ml) 15

Gas retention time in activated carbon (s) 2,45

Porisity of beads of alginat (ml) 35

Gas retention time in beads of alginat (s) 5,7

Porisity of Kaldnes K1 media (ml) 36

Gas retention time in Kaldnes K1 media (s) 5,9

Gas flow rate (ml/min) 366

Recirculation of substrate (ml/hour) 36

Surface loading (m3/m

2h) 70

Feature Value

Regenerated cellulose content (%) 45-55

Ash (%) 35-40

Particle size (mm) 2-3,5

Aggregate thickness (g/dm3) 250-300

water uptake at 25 C (%) 150-200

Special pore volume (cm3/g) 1.5-2

Special pore surface area (m2/g) 8-10

Swelling

increase in diameter 1,5 fold

increase in volume 3 fold

16

Figure 4: Set-up of the continuous biotrickling column reactor

17

(a)

(b)

Figure 5: H2S elimination by Thiobacillus thioparus immobilized on activated

carbon (a), amount of protein on the surface of the support (b)

Figure 6: The picture of the activated carbon plug in the column

0

10

20

30

40

50

60

70

80

90

0 50 100 150 200 250

Hy

dro

gen

-su

lph

ide c

on

cen

tra

tio

n [

pp

m]

Time [hour]

Vak

Thiobacillus thioparus

0

1

2

3

4

5

6

7

8

9

10

0 48 96 144 216

Pro

tein

co

ncen

tra

tio

n [

mg

/g s

up

po

rt]

Time [hour]

18

(a)

(b)

Figure 7: H2S removal by the bacteria immobilized on Mavicell B (a), amount of

protein on the surface of the support (b)

0

10

20

30

40

50

60

70

80

90

100

0

10

20

30

40

50

60

70

80

90

0 50 100 150 200 250

Rem

ov

al eff

icie

ncy

[%

]

Hy

dro

gen

-su

lph

ide

co

ncen

tra

tio

n [

pp

m]

Eli

min

ati

on

ca

pa

cit

y [g

H2

S /

m3

h1

]

Time [hour]

Thiobacillus thioparus

Control

Removal efficiency

Elimination capacity [g H2S /m3h]

0

1

2

3

4

5

6

7

8

9

10

0 72 120 216

Pro

tein

co

ncen

tra

tio

n [

mg

/g s

up

po

rt]

Time [hour]

19

(a)

(b)

Figure 8: H2S elimination by Thiobacillus thioparus immobilized on activated

carbon (a), amount of protein on the surface of the support (b)

0

10

20

30

40

50

60

70

80

90

100

0

10

20

30

40

50

60

70

80

90

100

0 50 100 150 200 250

Re

mo

val e

ffic

ien

cy [

%]

Hyd

roge

n s

ulp

hid

e c

on

cen

trat

ion

[p

pm

]El

imin

atio

n c

apac

ity

[g H

2S

/m3

h]

Time [hour]

Thiobacillus thioparus

Control

Removal efficiency

Elimination capacity [g H2S /m3h1]

0

2

4

6

8

10

12

0 48 120 192

Pro

tein

co

nce

ntr

atio

n [

mg/

g su

pp

ort

]

Time [Hour]